1. Field of the Invention
The present invention relates to a communication network system and, more particularly, to a quality-deteriorated area estimating device and the like for estimating the deteriorated area in terms of communication quality.
2. Description of the Related Art
Conventionally, communication-quality deteriorated area has been specified in the following manner at a high speed when there is deterioration in the communication quality of a communication network. That is, a number of monitoring devices are provided within the network for constantly monitoring the communication quality at the positions where the monitoring devices are provided and, when the communication quality is deteriorated, a section with quality-deterioration is estimated according to the position of the monitoring device where the quality is deteriorated. This method requires a great number of monitoring devices for improving the estimation accuracy of the position where the quality is deteriorated. Therefore, it is costly to be provided with the devices.
The quality-deteriorated area estimating method disclosed in Japanese Patent Unexamined Publication 2002-271392 (METHOD OF CONTROLLING VOICE QUALITY FOR EVERY CALL IN IP NETWORK) finds the route of the quality-deteriorated flow based on quality information of communication flow of a network user and network topology, and specifies the common route section of a plurality of quality-deteriorated flows for estimating it as a quality-deteriorated area.
For example, let us assume that in a network as shown in FIG. 24 in which routers (or packet switches) R1-R6 and terminals T1-T6 are connected, flow F1 and flow F2 of the user flow through common route from R1 to R2, and there is deterioration of the quality occurred in the route from R1 to R2. In that case, the flows F1 and F2 are quality-deteriorated. The conventional quality-deteriorated area estimating method estimates the route section from R1 to R2, which is used in common by the quality-deteriorated flows F1 and F2, as the quality-deteriorated area.
However, when the routes of the quality-deteriorated flows are similar and there are a great number of common route sections, this method estimates all of them as the quality-deteriorated areas. Thus, this method has poor estimation accuracy and, in addition, may estimate a wrong section as the quality-deteriorated area when there are a plurality of quality-deteriorated areas.
Specifically, for example, let us assume that in a network as shown in FIG. 25 in which routers (or packet switches) R1-R6 and terminals T1-T7 are connected, flow F1 and flow F2 of the user flow trough common route from R1 to R4, and there is deterioration of the quality occurred in the route from R1 to R2 as a part of the common route. In this case, the flows F1 and F2 are quality-deteriorated. However, the conventional quality-deteriorated area estimating method estimates all the communication-route sections from R1 to R2, R2 to R3, and R3 to R4, which are shared by the quality-deteriorated flows F1 and F2, as the sections suspected of having quality deterioration. This estimation is not the actual quality-deteriorated section and, in addition, wider sections than the actual quality-deteriorated section are estimated as the quality deteriorated sections. Therefore, it has low estimation accuracy.
As another example of the same network structure, let us assume that, when the flows F1-F4 flow by partially sharing the communication route sections as shown in FIG. 26, there is quality deterioration occurred in the partial communication route from R1 to R2 and R3 to R4. In that case, the qualities of the flows F1, F2, F3, and F4 are deteriorated. In this state, there is no route section that is shared by all the quality-deteriorated flows F1, F2, F3, and F4, and the route sections partially shared by two or more flows are the route sections from R1 to R2, R2 to R3, and R3 to R4. The conventional quality-deteriorated section estimating method estimates the common route as the quality-deteriorated section. Thus, it concludes to estimate that “there is no section suspected of quality deterioration” or “the sections from R1 to R2, R2 to R3, and R3 to R4 are suspected of quality deterioration”. However, the sections where the quality is actually deteriorated are the partially shared routes from R1 to R2 and R3 to R4. Therefore, the former estimation result is wrong and the latter estimation result is of poor accuracy.